Customer behavior on social media networks is continuing to draw the attention of businesses. In particular, businesses are increasingly interested in monitoring social media posts both to the business' social media networks as well as monitoring social media posts by known customers, regardless of whether such posts are on the business' social media network. As social media continues to evolve, so too must a business' strategy for dealing with social media interactions.
Existing methods of monitoring and responding to social media interactions are primarily concerned with detecting a posting, classifying whether the posting requires a response from the business, and, if so, preparing an appropriate response. Unfortunately, the classification process for identifying whether to respond to a particular posting still has significant problems. Current solutions attempt to classify a posting primarily based on the content of the posting. This can result in unnecessary contact center resources being deployed for certain posts that do not necessarily require a response or, more problematically, may result in the failure to deploy resources for certain posts that should warrant a response. For example, a user may make a neutral post about a product. Using current classification processes, a response would likely not be generated. However, if the user's previous responses have been always highly favorable, failure to respond could result in decreased customer loyalty or even in the loss of a loyal customer.